|
Ok goons, I have a pretty basic understanding of electronics and pretty much zero understanding of LEDs but I'm a plant biology major and I want to build a grow light that can blind God while simultaneously not burning my house down. Besides a heat sink and fan what's required to get a couple of these working? http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-NonDirectional/XLamp-CXA3070 Not necessarily the exact one I want because I'm not sure of the color spectrum but it's the brightest one I could find. What the gently caress is a LED driver and do I need one? Am I going to accidentally kill myself?(Yes.)
|
#
?
Oct 13, 2013 08:31
|
|
|
|
| # ? May 8, 2024 05:30 |
|
|
ghetto wormhole posted:Ok goons, I have a pretty basic understanding of electronics and pretty much zero understanding of LEDs but I'm a plant biology major and I want to build a grow light that can blind God while simultaneously not burning my house down. Besides a heat sink and fan what's required to get a couple of these working? http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-NonDirectional/XLamp-CXA3070 Not necessarily the exact one I want because I'm not sure of the color spectrum but it's the brightest one I could find. What the gently caress is a LED driver and do I need one? Am I going to accidentally kill myself?(Yes.) Unless for some reason you need to use a single >100W LED, you're probably better off with a small array of LEDs, as they're generally much easier to control and you'll have an easier time dealing with heat, plus you can mix LEDs to get the spectrum you want. If you're looking to make a grow light and you've got university journal access, you may want to look up Designing light-emitting diode arrays for uniform near-field irradiance by Moreno, Avendaņo-Alejo, and Tzonchev (2006). I've found it quite handy for building lab equipment. If the single giant LED is a must, you'll most likely want a circuit that will adjust the voltage to the appropriate level and regulate current. If you're starting with a voltage higher than your diode's rating, you'll want to look into a buck converter. If it's lower, you'll want to look into a boost converter. I don't have much experience with LEDs over 3 watts, so I'll leave the specifics to someone more knowledgeable.
|
|
#
?
Oct 13, 2013 08:50
|
|
|
I've been slowly converting everything I have to LED's with pre-made solutions like lightbulbs and things made mostly by me. I'd definitely check up on spectrum if your growing box is going to be for science purposes since "white" LED's can vary wildly. Also if you're going to be using one single giant LED in an enclosed area it's going to get very hot. Even the array of LED's will get hot but it's generally easier to manage if its spread out. Keep heat dissipation in mind when you're designing whatever this will be mounted in. LED's don't get as hot as other lamps, but the high-output ones still do get pretty hot if you're not careful. EDIT: Also, if you're using this for actual science, be aware that an LED's wavelength and luminescence both change a bit based on its temperature. Also, manufacturer specified values for color and luminescence are generally measured in a highly controlled laboratory and given for an unrealistic temperature like 25C which you'd never get in reality since they're usually fairly hot devices. This might gently caress up your experiments if you're doing some precision plant-growing, but otherwise isn't a big problem since it doesn't change that much. Shame Boy fucked around with this message at 10:20 on Oct 13, 2013 |
|
#
?
Oct 13, 2013 10:12
|
|
|
ghetto wormhole posted:Ok goons, I have a pretty basic understanding of electronics and pretty much zero understanding of LEDs but I'm a plant biology major and I want to build a grow light that can blind God while simultaneously not burning my house down. Besides a heat sink and fan what's required to get a couple of these working? http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-NonDirectional/XLamp-CXA3070 Not necessarily the exact one I want because I'm not sure of the color spectrum but it's the brightest one I could find. What the gently caress is a LED driver and do I need one? Am I going to accidentally kill myself?(Yes.) Would it be worth selecting specific red and green LEDs to target the spectral peaks that are best for photosynthesis, rather than going for a big gently caress off white one?
|
|
#
?
Oct 13, 2013 12:24
|
|
|
Gary Manifold posted:Would it be worth selecting specific red and green LEDs to target the spectral peaks that are best for photosynthesis, rather than going for a big gently caress off white one? That depends on what you are trying to grow and where it is in its lifecycle. I'd suggest getting into some research on what light temperatures are appropriate for propagation and plant growth vs. flowering and fruiting, because they are different. But if you're building this yourself you could make a single multi-led unit that you can switch as appropriate (which would be pretty cool for a small solution).
|
|
#
?
Oct 13, 2013 14:07
|
|
|
Gary Manifold posted:Would it be worth selecting specific red and green LEDs to target the spectral peaks that are best for photosynthesis, rather than going for a big gently caress off white one? You actually almost never want green light for plants, since they reflect most of it (hence why most plants look green). Most high-efficiency LED grow lights have a mixture of red, blue, and possibly UV. This is definitely something you want to be wary of doing on a work computer, but hands-down the best resources I've seen on custom fitting light spectra to plant growth (and different periods in the plant's life cycle) have been from stealth pot grow websites. Going with multiple LEDs instead of one big one makes it much easier to build modular systems, handle heat dissipation, and gives you more flexibility (say, running things at lower amperage in series or lower voltage in parallel). Going with a mix of different colored LEDs rather than white ones gives you a lot more flexibility in custom tailoring the output spectra to the needs of your plants.
|
|
#
?
Oct 13, 2013 19:33
|
|
|
ghetto wormhole posted:Ok goons, I have a pretty basic understanding of electronics and pretty much zero understanding of LEDs but I'm a plant biology major and I want to build a grow light that can blind God while simultaneously not burning my house down. Besides a heat sink and fan what's required to get a couple of these working? http://www.cree.com/LED-Components-and-Modules/Products/XLamp/Arrays-NonDirectional/XLamp-CXA3070 Not necessarily the exact one I want because I'm not sure of the color spectrum but it's the brightest one I could find. What the gently caress is a LED driver and do I need one? Am I going to accidentally kill myself?(Yes.) Other people have covered a few bits of important systems stuff, so I'll talk about the driver. If you put too put too little voltage across a bare LED, it doesn't turn on. If you put too much across it, an LED draws too much current and dies. What you want is something that limits the current going though the LED.* For small projects where you have a handful of weak LEDs, the usual current limiting device is a resistor. It's simple easy cheap and reliable. The disadvantages is that you need to resize it every time you change the number of LEDs. So if you're experimenting with different setups, or you're making equipment that needs drives however many LEDs the customer wants, you instead make a constant current supply, and set it to adjust its internal limiting resistor depending on how many LEDs there are. The other problem with these resistive circuits is that their efficiency is terrible. For example 1 small red led on a 5v circuit. If we take 20ma as our target current, and the LED having a voltage drop of 1.8v. To do this, we'd use a 160 ohm resistor. In this case, 36 milliwatts are consumed by the led, and 64 mw are consumed by the resistor. These numbers work out to 36% efficiency. Or worse really, as even the best LEDs are only 40% efficient. That means that the overall circuit is only about 14.4% efficient at turning electricity into light. When you're using it to drive a tiny indicator light that has a total power consumption of 100mw that's ok. You get 14.4mw of light and 85.6mw of heat. A typical light bulb replacement LED would be around 14watts or so. With a resistive current limiter, you'd be talking around 25watts of heat dissipated by the resistor per led array. Now this is a naive implementation but even with smarter ones resistive current limiters give you a heat problem pretty quickly. So you want something that isn't resistive. There are a couple of ways, but you've probably seen switchmode led drivers. These can be 80-90% efficient, so you can cut that 25watts of heat up there to 2.5w. Better with less naive layouts. To continue the comparison, these figures would give your light a 200w resistor. Now honestly this isn't really useful anymore; to get to this point in the design and need one that size so much would have gone wrong, (things like to make those figures fit, you'd have a 106.9v supply) but there you go. Given seeing as a good switchmode driver isn't the easiest thing to build, and that a resistive driver on any voltage that you'd have access to would also be a pretty great heater, I can say that you should absolutely by a driver for it. As an aside, that's an very powerful array. A 60W lightbulb produces around 800 lumens, that one is 9000 - 11000. *There is no perfect voltage. It is either too large or too small. Things like manufacturing variation, the shape of Vf vs If plots, that Vf varies over temperature get in the way. Some sources, most notably coin cells, have large internal resistances and naturally current limit though.
|
|
#
?
Oct 14, 2013 05:21
|
|
|
I was searching back through this thread to find some old posts, and came across this, which is probably the best thing ever posted here:longview posted:
|
|
#
?
Oct 16, 2013 17:35
|
|
|
Slanderer posted:I was searching back through this thread to find some old posts, and came across this, which is probably the best thing ever posted here: The funny thing for me was that I discovered that around 1% of the bags were empty, but a further 1% had two transistors. But if we're doing thread favourites I really liked the discrete transistor-ball opamp someone made ~1-2 years ago. longview fucked around with this message at 19:52 on Oct 16, 2013 |
|
#
?
Oct 16, 2013 19:44
|
|
|
longview posted:The funny thing for me was that I discovered that around 1% of the bags were empty, but a further 1% had two transistors. One of my friends did something like that. We had a school assignment to design a voltage buffer out of discrete BJTs and characterize it. This guy, being a somewhat lazy smartass took the LM741 schematic and built it on a breadboard. The professor was impressed, but made him redo it anyway.
|
|
#
?
Oct 17, 2013 02:13
|
|
|
Does anyone know if there's a specification or some common design for headsets with a 4 conductor 3.5mm jack that are used by non-apple devices? I used to have a nokia hands free kit that had a button on the cable that could be used to answer calls but I didn't keep it after I killed the earbuds so I'm not sure exactly what the button did. I know it used the outer sleeve as ground and I suspect the button worked by shorting the mic connector to ground but I'm not 100%. I have an iphone handsfree cable that I'm using as a set of earbuds by holding the button down which does short the mic connector to ground, but that's required because iphones seem to use the outer sleeve as the mic in and the second sleeve is the ground. (Compare a "typical" connector to an iphone headset connector). I'm looking to cannibalize a PC headset I have lying around to combine the separate 3 conductor headphone and microphone jacks into a single 4 conductor connector. I suppose I could just try it but it'd save me unsoldering these delicate wires if I get it wrong.
|
|
#
?
Oct 18, 2013 15:17
|
|
|
Ephphatha posted:Does anyone know if there's a specification or some common design for headsets with a 4 conductor 3.5mm jack that are used by non-apple devices? Non-Apple isn't specific enough. Surprisingly they aren't being a special little snowflake in this case. It's Nokia and Lenovo in the minority. 
|
|
#
?
Oct 18, 2013 17:00
|
|
|
May be a weird question but... how come there's a spark when I plug in a charger to a wall if it's plugged into the computer first, but not the other way around? Is there a way to design a circuit for charging lithium ion battery devices that allows for hot plugging on either end without spark?
|
|
#
?
Oct 21, 2013 22:54
|
|
|
I need to use solid state temperature control to get an aluminum block to a certain temperature. My idea was to use a Peltier module, with an LED current driver to control the temperature differential of the Peltier. However, a quick test of this shows it doesn't work. Parts I'm using: Peltier module (Digikey 926-1184-ND), BuckBlock (LEDSupply, A009-D-V-1400). After some googling I still can't get a good idea of why this doesn't work. Could someone explain it to me? Some searching says PWM should work, but there is some conflict on what frequency is needed. I've seen some say over 10 Hz is sufficient, some say over 500 Hz. I have a PWM controller that operates at 240 Hz, would that work? Are there any other simple methods of doing what I'm trying to do? I'm not skillful enough to create complicated circuits but I can handle simple ones.
|
|
#
?
Oct 21, 2013 23:10
|
|
|
"Doesn't work" is pretty vague. If the peltier simply isn't functioning, try hooking it directly to 12V to make sure it's actually functioning. Keep in mind your driver is fairly weak, and you'll probably see 20-25C delta at best.
|
|
#
?
Oct 21, 2013 23:40
|
|
|
If the temperature is above room temperature, you probably don't need a peltier, you just need a heater. Again, 'doesn't work' is vague. What you really need is a feedback loop from a thermocouple or temp sensor of some kind so that it can control the temperature. Sometimes this can just be simple negative feedback, but more often then not when it's a system with various thermal capacity/intertias, you need a PID loop. (There are PID libraries for arduino, but you have to learn how to tune it at least.)
|
|
#
?
Oct 22, 2013 01:04
|
|
|
Xovaan posted:May be a weird question but... how come there's a spark when I plug in a charger to a wall if it's plugged into the computer first, but not the other way around? Is there a way to design a circuit for charging lithium ion battery devices that allows for hot plugging on either end without spark? You mean on a laptop, right? My guess would be that the wall outlet is 110 or 220 V, while the laptop side is only 15-20 V. Why do you want to make it pluggable without sparking? If I had to guess at a way to do that, I might see about isolating the charger circuit from the AC input until a high impedance AC voltage checker indicated the voltage was present, then maybe relay it over? I don't know...this is out of my area.
|
|
#
?
Oct 22, 2013 01:09
|
|
|
Xovaan posted:May be a weird question but... how come there's a spark when I plug in a charger to a wall if it's plugged into the computer first, but not the other way around? Is there a way to design a circuit for charging lithium ion battery devices that allows for hot plugging on either end without spark? I assume this is a laptop charger, right? I'd imagine that the charger might have a larger inrush current if a load is present, for whatever reason. Here's my guess: When any pair of metal contacts come together, they may be some degree of "bouncing" as one of both of the contacts is still moving. When this is happening, it will appear that the switch is rapidly opening and closing. I don't know if this actually happens to a meaningful degree with AC plugs, but it does with relays and microswitches, for instance (those actually have the moving contact bounce off the stationary one a few times before it settles, essentially. with the AC plug, I'd imagine there might be some intermittent contact as the initial connection is made, due to the the plug not going in completely straight and the contacts not being perfect). If this happened, the initial "bounce" might provide enough energy to the charger to power up it's electronics. The electronics enable some FETs, and create a low-impedance path for AC current through a transformer or something. With that low impedance path present the breakdown voltage of the air in that small gap is exceeded and current can flow in an arc. Alternatively, if it was a weird charger that fed a voltage back into the charger electronics, the charger could be powered and enabled, creating a low-impedance path for arcing to occur right as the plug is initially connected. But that seems less likely. There are a few different ways to add inrush current limiting to a circuit you are designing, but it's harder for adding onto something that's already built. It may be possible to add an NTC thermistor to limit the inrush current. NTC thermistors are used as temperature sensors, since their resistance decreases with respect to temperature in a fairly predictable manner. In this case, they can be used on an AC or DC input to appear as a higher-impedance load when the device is disconnected (and therefore cold), and once the device is plugged in it will allow some current, and then quickly heat up and allow more and more current. There are a few drawbacks to this approach: -The NTC will dissipate a bunch of heat, making the charger less efficient, and making you deal with that heat -The charger must be able to handle the initial high-impedance AC supply -it will not work if the AC is unplugged and plugged immediately back in while in use, as the thermistor will remain hot (and therefore low impedance). I assumed that most supplies for laptops and desktop computers used NTCs for this already, so it's somewhat surprising yours doesn't. It probably does all the currently limiting in the electronics, which might get "tricked" by a bouncing switch. And perhaps the maker decided that it wasn't worth the extra parts and effort in order to prevent this, as (a) most people plug the charger into the wall first and (b) a tiny, single-shot arc inside an outlet isn't a big deal unless you're surrounded by combustible gasses. edit: I just tried this with the charger for my Dell laptop, and it definitely arcs when I replug the AC while the DC jack is still connected...but it also does this with the DC disconnected. However, I don't really care enough to unplug my charger for a while to let it cool down to see if there is an NTC element that will limit current once it cools down. Slanderer fucked around with this message at 01:48 on Oct 22, 2013 |
|
#
?
Oct 22, 2013 01:45
|
|
|
Rescue Toaster posted:If the temperature is above room temperature, you probably don't need a peltier, you just need a heater. I would need to reserve the option to cool the block instead of heating it, so that's why I selected a Peltier. The module definitely works, when connected to a 12 V 1.25 A power supply, the hot side gets too hot to touch within 10 seconds. However, using that 1400 mA driver results in no subjective change in temperature (I didn't measure the temperature, but I could not feel any temperature differential). Since the power supply puts out less current than the current driver limit, I had expected no change in function. I did also try a 350 mA current driver though, which also resulted in no apparent temperature differential. I do plan on using PID control using a thermocouple and the 10 V external dimming control on the current driver to reach a set-point temperature. I have experience with PID tuning and have already used PID control with success in various other systems.
|
|
#
?
Oct 22, 2013 02:06
|
|
|
Hidden Under a Hat posted:I would need to reserve the option to cool the block instead of heating it, so that's why I selected a Peltier. Perhaps the led driver expects a diode as it's load. You could try placing something like a 1N5817 in series with the peltier. Or stick an oscilloscope on the output, see if it's generating a AC output. A couple of notes: 1. You could use a more common 1n4001, but even 200hz is a bit fast for one. 2. A full 1.25A is too much current. Turn it like half on, and see if it works. Or you could use a 1N5820, which is a 3A version. Or try building the circuit here, see if it gives you a variable control. Replace the motor with the peltier, and since the a peltier won't generate inductive kicks you don't need the flyback diode either. http://forum.allaboutcircuits.com/showthread.php?t=73323
|
|
#
?
Oct 22, 2013 05:34
|
|
|
Thanks for the replies! Another question: How does a battery *properly* charge while in use? I know how a car charging circuit works, but I'm confident there's more to it than "power supply = alternator", right?
|
|
#
?
Oct 22, 2013 06:52
|
|
|
Depends on the battery chemistry. Some (lead-acids, nicad, NiMH) you can just stick on a constant-current charger at some low value forever and they won't be harmed. For instance, you can happily maintain a car battery on a 12v 250mA wall wart with stripped wires. Other kinds of batteries (mostly lithium chemistries) are really bad at handling overcharge and need to be carefully monitored when they reach the end of the cycle. The best smart chargers out there usually have a three step process: 1) Charge at constant current equal to ~1C (1x the capacity per hour; a 2.2Ah battery charging at 1C draws 2.2 amps) until the battery voltage reaches its peak. For a completely dead battery this will be around an hour. 2) Switch to constant voltage mode and maintain the battery at peak voltage, decreasing the current as needed to prevent the voltage from rising. How long this takes depends on self-discharge, ion mobility within the battery, and a whole pile of other things. 3) Depending on the chemistry, either continue reducing the current until it's a trickle charge that matches the self-discharge rate (best practice for lead-acid, etc); or continue charging until the charge current is about 2% of the peak current, at which point you call it charged and terminate the cycle to avoid starting fires. I don't know of any chargers that are smart enough to figure this out on their own so you usually have to at least tell them whether it's a lithium battery or not. When your iPhone does its fast charge thing that brings it from dead to 80% in an hour, it's basically just doing step (1) and maybe even boosting above 1C depending on how confident the Apple engineers were in their battery supplier's quality control. The remaining hour and a half it takes to get to 100% is steps (2) and (3). I've built constant-current and constant-voltage power supplies before, but I'm kind of foggy on the schematics right now so if you're asking how the electronics specifically work someone else can probably answer that better than I.
|
|
#
?
Oct 22, 2013 07:07
|
|
|
Xovaan posted:Thanks for the replies! Another question: How does a battery *properly* charge while in use? I know how a car charging circuit works, but I'm confident there's more to it than "power supply = alternator", right? Here is where you insert the sad trumpet noise, I think. As far as I know, alternators are usually just the three phase motor connected to the battery through a rectifier circuit (which is just 6 heavy duty diodes). The whole thing is stupid simple and works based on the simple fact the three components are pretty bombproof.
|
|
#
?
Oct 22, 2013 07:13
|
|
|
There's generally also a regulator in there somewhere, since the alternator speed depends on engine speed (in my motorcycle, the rotor is literally mounted on the crankshaft) so when you rev it up high the voltage goes sky-high too. Without the regulator you can easily see 30-40 volts at the rectifier, plenty high to boil off all your electrolyte. e: though I think more modern vehicles that have more space use energized rotor windings, so that you can vary the magnetic field and hence the voltage more efficiently than a chop-off regulator Sagebrush fucked around with this message at 07:26 on Oct 22, 2013 |
|
#
?
Oct 22, 2013 07:22
|
|
|
Xovaan posted:Thanks for the replies! Another question: How does a battery *properly* charge while in use? I know how a car charging circuit works, but I'm confident there's more to it than "power supply = alternator", right? As other people have said, charging lead acid batteries is extremely crude and simple, which is why they have been in use so long. Other batteries (NiMH and LiIon) are somewhat more complicated to charge, and less resilient.
|
|
#
?
Oct 22, 2013 17:51
|
|
|
Sagebrush posted:There's generally also a regulator in there somewhere, since the alternator speed depends on engine speed (in my motorcycle, the rotor is literally mounted on the crankshaft) so when you rev it up high the voltage goes sky-high too. Without the regulator you can easily see 30-40 volts at the rectifier, plenty high to boil off all your electrolyte. On the vehicles I've worked on, there is no regulator between the alternator and the battery. I have never seen one elsewhere, but I haven't had to do much electrical work, mostly transmission work (gently caress ford forever). If there is a regulator on the rest of the system, it has got to be pretty piss-poor. The 12V on a car is usually noisy as gently caress and ranges between 10-11V and 14-15V and close 0V during cranking. As far as I understand, the battery provides most of the regulation as it can absorb a lot of current without its voltage changing much. EDIT: Now that I think about it, at the very least it's likely the accessory stuff in the dash is regulated on newer cars. I know it wasn't on my truck, the cig lighter was a direct connection to the battery. Delta-Wye fucked around with this message at 19:24 on Oct 22, 2013 |
|
#
?
Oct 22, 2013 19:21
|
|
|
Delta-Wye posted:On the vehicles I've worked on, there is no regulator between the alternator and the battery. I have never seen one elsewhere, but I haven't had to do much electrical work, mostly transmission work (gently caress ford forever). If there is a regulator on the rest of the system, it has got to be pretty piss-poor. The 12V on a car is usually noisy as gently caress and ranges between 10-11V and 14-15V and close 0V during cranking. As far as I understand, the battery provides most of the regulation as it can absorb a lot of current without its voltage changing much. A lot of alternators have integrated regulators. This is why the Delco 10SI and 12SI are popular among tuners, classic car enthusiasts, and hobbyists-- Ground connection via the chassis, one wire to the battery, and you're in business.
|
|
#
?
Oct 22, 2013 20:25
|
|
|
Awesome answers. Thanks for the replies, guys! Tons of useful/interesting info.  Next step is building my own Pass Class A speaker amplifier 
|
|
#
?
Oct 23, 2013 06:09
|
|
|
I am trying to fix several radio control cars. They are very tiny, and fairly straight forward in terms of repairs - the old Ni-Cd batteries had never been charged, they leaked (blue power inside) which has corroded the solder joints and all the wires fell off when I cracked it open to see why it wouldnt work. The steering is done using two electro magnets which are just coils of fine wire. Unfortunately I keep breaking the original ones trying to solder them to the PCB. Is there a method for winding coils correctly? I figure the only way I can fix these is if I make new coils, as the wire is snapped on the inner diameter and therefore impossible to extract and unwind further. Furthermore, is there any rules on how many winds? ie will more/less winds and/or gauge of wire affect the power of the coil? to summarise: how do you accurately wind coils to form electro magnets, is there a tool? how do you wind them tightly and get them to stay coiled? does the composition of the coil (winds/gauge) affect performance?
|
|
#
?
Oct 24, 2013 14:13
|
|
|
Hello blue smoke thread. I made a thing. Alcohol swab: https://www.youtube.com/watch?v=veQYlWzNASQ Brine swab (holy el oh el, batman!): https://www.youtube.com/watch?v=y5sAvHxxv4s Finally, spraying tap water: https://www.youtube.com/watch?v=0iHljGYIfL0 okay, fine, a prototype thing
|
|
#
?
Oct 24, 2013 19:04
|
|
|
Laserface posted:to summarise: Coils are wound in the factory by spinning the coil while a spool of wire moves around it. It's sort of like a weaving loom: https://www.youtube.com/watch?v=6lvm3FGTHSI You can do stuff like this yourself, although more slowly, with a power drill and some cleverness. The more turns in a coil, the more powerful the magnetic field. You can also increase the field by increasing the current, or by making the coil itself smaller (narrower radius). The gauge of the wire doesn't directly affect the field strength, but it does affect indirectly through the number of turns you can fit into a given space (larger wire = fewer turns) and the amount of current the coil as a whole can carry (larger wire = more current). I don't remember the laws exactly but there are lots of quadratic functions so the tradeoffs get fairly complicated.
|
|
#
?
Oct 25, 2013 00:07
|
|
|
Hi everybody! Wow, haven't posted in here for ages. Okay, theory question time. You know how for a lossless transmission line you can analyze it by breaking it down into lumped elements, deriving the telegrapher's equations, and finally coming up with this beauty: Z(l) = Z0 * [ZL + jZ0tan(x) / Z0 + jZLtan(x)] My question, simply enough, is is there a version of that for a (uniform) lossy line? I've been searching on this extensively and I haven't come across anything. Is there really no closed-form equation for calculating the input impedance of a lossy tline?
|
|
#
?
Oct 25, 2013 15:52
|
|
|
Cyril Sneer posted:Hi everybody! Wow, haven't posted in here for ages. I would need to check somewhere before I write something wrong, but any good introduction to S-Parameters would describe this. Jup. checked it. Same formula only you change \beta for -j\gamma . Marcuvitz, page 26. 
VictualSquid fucked around with this message at 17:16 on Oct 25, 2013 |
|
#
?
Oct 25, 2013 16:38
|
|
|
So in my everlasting quest to be a nerd of ever-greater magnitude, I'm thinking about about making bases for D&D minis that will power an LED for various purposes. The first iteration of this involved a standing stone with a watch battery and an LED jammed up inside it. It looked cool, but those batteries run out and it's not terribly practical for any other mini. So then for a while I looked at making some sort of circuit board that would sit in a regular flat mini base. Had to fit in a circle a little under 1" diameter, and not be much taller than the circuit board itself. Problem there was that I'd either have to use a teeny tiny battery which would mean less brightness from the LED, or I'd have to use a bigger battery which would necessitate a taller base. Also, I'd need a good way to turn it on/off. In the end, I wasn't excited about that plan. The third option I've come up with involves wireless power. At the moment, I'm thinking this is the best plan. I really don't need much power at all, just enough to drive an LED of moderate brightness, so low efficiency isn't going to be a huge problem. Neither would turning the mini off be a problem: just remove it from the pad. And I'm hoping the space constraints won't be a problem, either: with the wireless power, all I really need (I think) is the PCB with a coiled loop trace on it (probably on both sides to get more turns, maybe even a 4-layer for even more turns), a capacitor, and the LED itself. So wireless power gurus: am I way off base here? Wireless power is mostly voodoo in my book, at least as far as tuning the correct values of capacitance, inductance, voltage, current, frequency, etc. I get the basic premise, but I'm looking for a sanity check before I start trying to make things happen. Here's what I'm thinking: a circuit board that looks something like this:  That would be the "receiver," with 24 turns using the top and bottom. In the middle of the coil, I would have my LED and capacitor. Then, for the "transmitter" portion (I don't know what a more appropriate term would be here...inductor and inductee?  ) I'd have a similar 24 turns powered by some frequency of square wave at the desired voltage (my understanding is that with the same number of turns, the voltage in both coils will match.) Since this isn't for trying to quickly charge a battery or anything, really just getting a few mA for an LED, getting it perfect shouldn't be a huge concern and even a large amount of loss should be acceptable. I think.Also, is 24 turns enough? With the minimum trace width/spacing at, say, OSH Park, that's as much as I can fit in the space I have...unless I went to a four layer board, then I could get up around 50 turns, assuming they'll do blind vias. So yeah, anyone want to coach me along a little bit here?  e: For reference, here was the first iteration: http://forums.somethingawful.com/showthread.php?threadid=2734977&pagenumber=131&perpage=40#post395804888 Bad Munki fucked around with this message at 20:10 on Oct 25, 2013 |
|
#
?
Oct 25, 2013 20:05
|
|
|
I would use one of these tiny LiPo batteries personally. Since you can recharge them, it'll pay for itself in cr2032 coin cells relatively fast. I can't help you with the wireless option but I am interested to hear what other people come up with.
|
|
#
?
Oct 25, 2013 20:15
|
|
|
At almost .2" thick, that's way too bulky, unfortunately. My space constraints are a cylinder 0.9" in diameter and 0.105" thick. It's pretty cozy. The LED can be taller, since it'll poke up through the top of the base anyhow, but the rest of it has to be pretty shallow. But yeah, I feel like getting 5-10mA out of wireless like that should be do-able. The only other thing is that I want the transmitter pad to be large or, preferably, modular, so I could have like a grid of transmitters covering an arbitrary surface and assemble them to cover whatever surface area I need. I'm just not sure if I can arrange them in a grid like that and have it still work. Bad Munki fucked around with this message at 20:27 on Oct 25, 2013 |
|
#
?
Oct 25, 2013 20:25
|
|
|
Hey, thread, I just finished a thing. https://www.youtube.com/watch?v=DXPymiiYudA It's a 25-key MIDI keyboard made out of touch-sensitive nickels and pennies. I did a full build write-up if anyone cares.
|
|
#
?
Oct 26, 2013 22:23
|
|
|
That's pretty rad. I've been leaning towards doing a MIDI project like that shortly. What made you choose classic MIDI over USB if you're going straight to the computer? What did you use to connect MIDI to the computer?
|
|
#
?
Oct 27, 2013 00:16
|
|
|
So I played with this wireless power thing a little bit. I don't know what I was smoking before, but my sizes were waaaaay off. Anyhow, here's an updated receiver module. Remember, that outer circle is only 0.9" in diameter. The real action is in the middle: dead center you can see the footprint for an 0603 package, that's a capacitor. On either side of that are a couple through holes and pads. Those are for a standard 5mm LED. My plan, to make the most use of the available space, is to have the LED sitting right on top of the capacitor, straddling it like some sort of capacitor rodeo. It shouldn't matter that the LED will stand a little higher because of the way this thing will be mounted. Anyhow, if my understanding is correct, the power induced in the coil will be AC, but will be oscillating super duper fast, meaning the LED will have like a 50% duty cycle but at something like 150khz at least so that should be okay (150khz is just the value from the plans I'm working off of...sounds like it can be much, much higher. That's okay too, I think.) I guess I could add some stuff to convert it to DC but I want to keep this as absolutely dirt simple as is possible. Bad Munki fucked around with this message at 03:52 on Oct 27, 2013 |
|
#
?
Oct 27, 2013 03:38
|
|
|
|
| # ? May 8, 2024 05:30 |
|
|
Delta-Wye posted:That's pretty rad. I've been leaning towards doing a MIDI project like that shortly. What made you choose classic MIDI over USB if you're going straight to the computer? What did you use to connect MIDI to the computer? I did classic MIDI for simplicity/cheapness. I'm using a USB MIDI converter to the PC, but I didn't want to integrate the converter it because it cost a few bucks and seemed worth reusing. Also I may want to the Nickelphone with other classic MIDI gear in the future. Stabby McDamage fucked around with this message at 04:40 on Oct 28, 2013 |
|
#
?
Oct 27, 2013 03:52
|
|